420 likes | 525 Views
Business Intelligence. Lecture 15. Outlines. Why integrate BI into your smart client application? Integrating Data Mining into your application Integrating Reporting into your application. What Is Business Intelligence?.
E N D
Business Intelligence Lecture 15
Outlines • Why integrate BI into your smart client application? • Integrating Data Mining into your application • Integrating Reporting into your application
What Is Business Intelligence? • “Business intelligence (BI) is a broad category of business processes, application software and other technologies for gathering, storing, analyzing, and providing access to data to help users make better business decisions” • Why integrate BI into your app? • BI makes applications smarter! • Your customers get more value from their data • Big payoff with small amounts of coding
What Is SQL Server Data Mining? • Data Mining platform for analysis and deployment of data mining solutions • Nine algorithms developed in conjunction with Microsoft Research • Tools for the creation and comprehension of models • Familiar API’s and query language for embedding data mining into intelligent applications
Data To Predict Training Data Mining Model Mining Model Mining Model What Does Data Mining Do?Illustrated DB dataClient dataApplication data DB data Client data Application data “Just one row” DMEngine DM Engine Predicted Data
Intelligent Applications • Make decisions without coding • Data Mining algorithms learn business rules directly from the data, freeing you from trying discover and code them yourselves • Customized for each client • Data Mining learns the rules from the client’s data – resulting in logic that is automatically specialized for each individual client • Automatically update themselves • As your client’s business changes, so do the factors that impact their business. Data Mining allows your application logic to be automatically updated through a simple processing step. Applications do not need to be rewritten, recompiled or redeployed, and are always online – even during processing
Reporting & Analysis Devices Business Analysis EnterpriseReporting Performance Management Information Delivery SourceSystems Client Portal EnterpriseReporting Data Marts ERP Business Performance Management Familiar,PowerfulBI Tools CRM Third Party Applications LOB EnterpriseETL DataWarehouse Data Analysis(OLAP, DataMining) Data Visualization Geospatial Project management Integrated Enterprise BI Platform
Integrate Analyze Report Data acquisition from source systems and integration Data transformation and synthesis Data enrichment, with business logic, hierarchical views Data discovery via data mining Data presentation and distribution Data access for the masses SQL Server 2005 SQL Server 2005 launches November 7, 2005
Alerts & escalation Call centre data: semi structured Data mining ETL Text Mining Staging Legacy data: binary files Staging Warehouse ETL Hand coding Staging Cleansing & ETL Reports Application database ETL Mobile data Example: before Integration Services • Integration and warehousing require separate, staged, operations. • Preparation of data requires different, often incompatible, tools – difficult to audit. • Reporting and escalation is a slow process, delaying smart responses – sequential operations. • Heavy data volumes make this scenario increasingly unworkable.
Alerts & escalation Mobile data Text mining components Call centre: semi-structured data Merges Data cleansing components Data mining components Standard sources Custom source Warehouse Legacy data: binary files Reports SQL Server Integration Services Application database Example: with Integration Services • Integration and warehousing are a seamless, manageable, operation. • Sourced, prepare and load data in a single, auditable process. • Reporting and escalation can be parallelized with the warehouse load. • Scales to handle heavy and complex data requirements.
Enterprise ETL platform • High performance • High scale • More trustworthy and reliable • Best in class usability • Rich development environment • Source control • Visual debugging of control flow and data • Great range of transforms out-of-the-box • Highly extensible • Custom tasks • Custom enumerations • Custom transformations • Custom data sources
Datamart Reporting Tool (3) Datamart Reporting Tool (2) Enterprise BI Today Data Source Data Model Tool MOLAP Online Analytical Processing (OLAP) Browser (1) MOLAP OLAP Browser (2) data warehouse (DW) Reporting Tool (1)
OLAP Browser (1) OLAP versus Reporting OLAP Browser (2) Datamart Datamart Duplicate Data Duplicate Models Enterprise BI – A Messy Reality Data Source Data Model Tool MOLAP MOLAP Reporting Tool (3) Reporting Tool (2) DW Reporting Tool (1)
OLAP Browser (1) OLAP Browser (2) Datamart Datamart Enterprise BI – With A UDM Data Source Data Model Tool MOLAP MOLAP UDM Reporting Tool (2) Reporting Tool (1) DW BI Applications
OLAP Browser (1) OLAP Browser (2) Datamart Datamart Enterprise BI – With A UDM Data Source Data Model Tool MOLAP MOLAP UDM Reporting Tool (2) Reporting Tool (1) DW BI Applications A single dimensional model for all OLAP analysis and Relational reporting needs
Analysis Services • Unified Dimensional Model • Integrating relational and OLAP views • Advanced Business Intelligence • KPIs, MDX scripts, translations, currency… • Web services • Native XML/A • Critical Server Abilities • Pro-active caching • Bringing the best of MOLAP to ROLAP
5 new algorithms Graphical tools/wizards 12 embeddable viewers Tightly integrated with AS, DTS, Reporting Integration with Web/Office apps Business Knowledge Data Mining OLAP Relative Business Value Reports (Adhoc) Reports (Static) Easy Difficult Usability Value of Data Mining SQL Server 2005
Decision Trees Clustering Time Series Naïve Bayes Sequence Clustering Association Neural Net New Set of Algorithms Introduced in SQL Server 2000
Mining Model Algorithms • Microsoft Decision Trees • Microsoft Clustering • Microsoft Naïve Bayes • Microsoft Time Series • Microsoft Association • Microsoft Sequence Clustering • Microsoft Neural Network • Microsoft Linear Regression • Microsoft Logistic Regression
BI Dev Studio (Visual Studio) Your Application OLE DB/ ADOMD/ XMLA App Data Deploy Analysis Services Server Mining Model Data Source Data Mining Algorithm Server Mining Architecture
Programmatically Creating Models • Analysis Management Objects • Object model for administrative tasks • Data Mining Extensions (DMX) CREATE MINING MODELTargetMailDT(CustID LONG KEY,Gender TEXT DISCRETE,CommuteDist TEXT DISCRETE,Education LONGCONTINUOUS,…BikeBuyer LONG DISCRETE PREDICT)USING Microsoft_Decision_Trees INSERT INTOTargetMailDT(CustID, Gender, CommuteDist, Education, …, BikeBuyer)OPENQUERY([My Data Source], ‘SELECT CustID, Gender, ComDist, Education, … BikeBuyer’)
Programmatically Creating Models • Analysis Management Objects • Object model for administrative tasks • Data Mining Extensions (DMX) • Dynamic mining using client data CREATE SESSION MINING MODELTargetMailDT(CustID LONG KEY,Gender TEXT DISCRETE,CommuteDist TEXT DISCRETE,Education LONGCONTINUOUS,…BikeBuyer LONG DISCRETE PREDICT)USING Microsoft_Decision_Trees INSERT INTOTargetMailDT(CustID, Gender, CommuteDist, Education, …, BikeBuyer)@InputRowset
Your Application App Data Data Source Retrieve Data Local Mining Architecture ADOMD.Net/OLE DB Local Analysis Services (msmdlocal) Mining Model Model File Decision Tree/Clustering algorithms
Why Integrate Reporting Services Into Your App? • Process data efficiently: filtering, sorting, grouping, aggregations, etc. • Present data in a variety of ways: lists, tables,charts, matrices • Add visual appeal to data: fonts, colors, border styles, background images, etc. • Enable interactive reports: collapsible sections, document map, bookmarks, interactive sorting, etc. • Support printing and print preview • Export to multiple formats (Microsoft Office Excel, PDF, XML) • Integrate with managed report server environment: security, caching, scheduling, delivery, etc. Bottom Line: More functionality, less code
Reporting Services In Visual Studio 2005 • Microsoft Visual Studio 2005 includesReporting Services functionality in standard editions and above • Window Forms and ASP.NET report viewer controls make it easy to embed Reporting Services functionality in applications • Integrated report designer for designing and embedding reports in language projects • Freely redistributable with no run-time restrictions • Report Server mode requires a SQL Server 2005 license
ReportDesigner (Visual Studio) Report Viewer(Remote Mode) Publish Report Report Server Web Services API DataSource Report Processing and Rendering SQL Server Catalog (Metadata) Report Server Mode Your Application
Your Application Report Viewer(Local Mode) Report Processing ReportDesigner (Visual Studio) DataSource Report Definition Report Data Embed Report Retrieve Data Local Mode
Designing In Local Mode • Report creation is integrated with Windows or Web language project (Add New Item…) • Integrates with Visual Studio data tools • Build report from project data sources (XSD) • Supports Data Sources window • Supports Data Source Configuration Wizard • Supports TableAdapter Configuration Wizard • Full Interactivity • Expand/Collapse sections • Drillthrough reports, including drillthrough from charts • Interactive sorting
Reporting Services • Introduced with SQL Server 2000 • Open, extensible enterprise reporting solution • Report authoring, management, delivery • Office System integration • VS.NET development environment • SQL Server 2005 enhancements • Integration with AS, DTS, management tools • Developer enhancements • Improved report interactivity • Rich end-user reporting
Rich Client Printing Multi-Valued Parameters Date Picker Parameters Sorting
Export to Excel, PDF E-mail Delivery
Report Builder • A new ad-hoc report design tool for Reporting Services • Targeted at business users who want to find and share answers to interesting questions • Driven from a business model of the data so users do not need to understand the underlying data structures • Fully integrated with Reporting Services and delivered in SQL Server 2005
Report Builder Client • Users download and launch the client from Report Manager (new or existing report) • Finished reports can be saved on the server
Report Builder Model Designer • For editing models used by Report Builder • Auto-generation of models from data sources
Application Examples • Healthcare • Builds models predicting health degradation in citizens • Sends reports to physicians of top patients likely to get sick • Alert Governments with predictable decease • Marketing • Sales Forecast • Targeting client • Traffic • Real-time predictive traffic info delivered to web and mobile devices • Everything…………………………….
Resources • msdn.microsoft.com/sql • SQL Server Data Mining • www.SQLServerDataMining.com • SQL Server Forums • http://forums.microsoft.com/msdn